One photo in · Same face, same room · Nothing to install
A bump on your own photo, in your own clothes, in your own front room — four months or nine, you choose.
A real run on this page's engine. Drag the divider and watch the ribbing on the top: the horizontal lines curve over the shape instead of staying straight, and the free hand moves. Every output is AI-generated, so your result varies with your photo — treat this as what the tool does rather than a result you will get back identically.


Drag to compare before and after
It reshapes the body in one photograph to carry a bump and leaves the face, the hair, the clothing and the room untouched — a body edit rather than a new picture of a different person.
Most versions of this effect fail in the same place: they draw a rounded shape on the stomach and stop. The result reads as a sticker because clothing is the tell. Ribbed knit runs in horizontal lines, and those lines have to bend over a curve; a waistband has to sit lower and looser; a shirt has to pull at the button that is now under tension.
Get that wrong and every viewer knows something is off even if they cannot name it, which is why this edit holds the fabric, the seams and the folds as part of the change rather than painting over them.
The second tell is posture. A body carrying weight in front stands differently — the lower back arches a little, the shoulders drift back, a hand very often goes to the top or the underside. The pass runs on Nano Banana 2 and shifts the stance along with the silhouette, then puts everything else back exactly as it was: same face, same expression, same hairstyle, same clothes in the same colour, same light coming through the same window.
Side-on or three-quarter photos read best, because a bump seen straight from the front is mostly hidden by the body itself.
Creative engine
Upload a photo, say how far along, and the clothing and stance change with the silhouette.
Ribs, seams and folds bend over the shape rather than sitting flat across it, which is what separates a body edit from a pasted-on shape.
Face, hair, expression, clothing colour, lighting and background come through the edit unchanged.
Other passes that reshape a real body rather than restyling it: build in the AI muscle generator, a different stance entirely in the AI pose changer, or a look at the face that comes next in the AI baby generator.
From an ordinary photo to a pregnant portrait that holds together.
Three-quarter or profile, fitted clothing rather than a loose jumper, whole torso in frame. A straight-on shot hides the very curve you are adding.
Four months reads as barely there, seven as clearly showing, nine as low and heavy. Naming the stage is more reliable than asking for a bump and hoping.
Look at the seams and the waistband first — that is where a weak result shows. Rerun with a fitted top if the clothing did not react.
Announcements, catching up on months nobody photographed, and stories that need a character to be pregnant.
A picture that says it without a caption, made from a photo you already like rather than a shoot you have to organise while exhausted.
Second pregnancies rarely get the camera the first one did. If the photo exists but the bump was never shot properly, this fills the gap in the album.
Writers, illustrators and video makers who need the same person at a different point in a story, the same way they would use the AI pose changer for a different stance.
Common questions about adding a bump to a photo with Kavel.
It works, but you lose the best evidence. A fitted knit or a t-shirt shows the shape through tension in the fabric; an oversized jumper hides it, so the result relies entirely on the silhouette and reads weaker. If you have a choice of source photos, pick the fitted one.
It responds to stages rather than dates — early, showing, late, twins — because that is what a photograph can actually convey. Nobody can tell twenty-six weeks from twenty-eight in a picture, and asking for a precise week gets you the model's average of that whole range anyway.
No. The edit is confined to the torso, the clothing over it and the stance that carries it. Face, expression, hairstyle, skin and everything behind you are supposed to come through untouched, and a result that changes your face is a failed run worth rerunning.
Yes — say which person the change applies to and what the other one is doing, such as resting a hand on the bump. Naming both roles keeps the edit from spreading to the wrong body, which is the usual failure on two-person sources.
Only with their agreement. Making an image of a real person that implies something untrue about their body is the kind of thing that lands badly however it was meant, and the content policy covers what you generate here. On yourself, or with the person in the photo asking for it, none of that applies.
Usually the source was cropped just below the chest, so the model had very little torso to work with and put the curve where the frame allowed. Use a photo that includes the whole torso down to the hips and it lands where it should.
The estimate is shown before you generate and changes with the resolution you choose, so nothing is a surprise afterwards. Failed runs return to your balance unless they broke the content policy, and a new account has enough to try a couple of stages.
Generate at the largest format offered rather than enlarging afterwards, since upscaling never recovers detail that was never rendered. Keep the original file — a version that has been through a messaging app has already been re-compressed.
One credit pool covers Nano Banana images and Seedance and Veo video. The cost shows before every run, and failed jobs are not charged. Use a subscription for ongoing work, or a one-time pack when you just need to top up.
What's included
Secure checkout by Stripe. Card details never touch our servers.
What's included
Secure checkout by Stripe. Card details never touch our servers.
What's included
Secure checkout by Stripe. Card details never touch our servers.
What's included
Secure checkout by Stripe. Card details never touch our servers.
One-time top-ups — buy extra credits any time you run low.
What's included
Secure checkout by Stripe. Card details never touch our servers.
What's included
Secure checkout by Stripe. Card details never touch our servers.
What's included
Secure checkout by Stripe. Card details never touch our servers.
Charges appear as “KAVEL AI” on your card statement. You can cancel any time from Settings → Billing; cancellation takes effect at the end of the period you already paid for.
Powered by
Use one balance across every supported image and video model — eleven of them today, listed below — and check the credit cost before each request.
Video models
Image models
Model credit guide
You're only charged for successful generations. The exact estimate in the generator varies by model, length, resolution, audio, and number of images.
| Type | Model | Credit cost |
|---|---|---|
| Video | Seedance 2.5 | The long-take tier, priced per second: 5s 480p ≈ 211 credits, 5s 720p ≈ 473. A full 30s take runs ≈ 1,260 at 480p and ≈ 2,835 at 720p. Supplying a reference clip lowers the per-second rate. |
| Video | Seedance 2.0 | 5s 720p image-to-video ≈ 188 credits; text-to-video ≈ 308 credits. Scales with resolution and length. |
| Video | Seedance 2 Fast | Faster and lower cost. 5s 720p text-to-video ≈ 248 credits. |
| Video | Seedance 2 Mini | The cheapest Seedance tier. 5s 720p ≈ 154 credits, 5s 480p ≈ 72 credits. |
| Video | Seedance 1.5 Pro | Audio doubles the rate. 5s 720p ≈ 27 credits silent, ≈ 53 with audio; 1080p ≈ 57 and ≈ 113. |
| Video | Veo 3.1 | Billed per video, not per second (Lite tier). About 45 credits at 720p and 53 at 1080p. |
| Video | Kling 3.0 | Audio raises the rate. 5s 720p ≈ 105 credits silent, ≈ 150 with audio; 1080p ≈ 135 and ≈ 203. |
| Video | MiniMax H3 | Fixed 2K, no resolution ladder. Priced per second — a 5s clip ≈ 158 credits. |
| Image | Nano Banana 2 | Generate or edit from text and images. About 8 credits per 1K image, 12 at 2K, 18 at 4K. |
| Image | Nano Banana Pro | Consistent run times across generations. About 12 credits per 1K or 2K image, 21 at 4K. |
| Image | Nano Banana 2 Lite | Faster, simpler variant. A flat 8 credits per image at every resolution. Start here to test. |
| Image | GPT Image 2 | The lowest-cost image model here. About 3 credits per 1K image, 6 at 2K, 12 at 4K. |
| Image | Seedream 5.0 Lite | Flat pricing — about 8 credits per image whether you generate or edit. |
One photo in, a maternity shot out — your face, your clothes, your room, with the fabric and the stance doing what they should.